2.3.6.0 CONCLUSION The features of multiple regression analyses and multicollinearity introduced in this unit are extension of unit 2. Here, we pointed out some of the complications arising from the introduction of several descriptive variables. In the discussions, we explained that when we go beyond the two-variable model and consider multiple regression models we add the assumption that there is no perfect multicollinearity (assumption 10 of CLRM). That is, there are no perfect linear relationships among the descriptive variables when two or more of these variables move together and difficult to determine their separate influences.
